Generally, due to increase in the complexity of virtual environment within an information technology infrastructure these days, it requires significant emphasis on evolving further a real time measurement technology with predictions to quantitatively depict the health of the virtual environment. Existing techniques captures a system utilization metric to ascertain the health of a virtual environment.
System utilization is not just the most commonly available metric across all operating systems; it is most easily available to all user levels. Whether it is an application developer or a system administrator, they can look at the central processing unit (CPU) utilization on any machine and get a good understanding of the current load on the system. However, the main challenge with a CPU utilization metric is that it does not reflect the contentions existing in the system. Whenever two processes contend for the same resource, e.g. CPU cycles itself, the operating system does a time slicing and assigns CPU cycles in a specific ratio to each process.
Similarly, in the virtualized environments, virtual machines are considered as user level processes. Whenever a process of higher priority enters the system, e.g. a real time process, the hypervisor has to pause the entitlement of a virtual machine and assign the CPU cycles to the real time process. In such cases, there is a high chance of virtual machine getting into a contention with respect to CPU cycles needed. In such cases, the system utilization shows very low, but at the same time the system is in a deadlock state. This essentially means that the CPU cycles are available but can't be used for any useful purpose by any process. Repeated happening of this event indicates bad health of the system; however it will not be caught by monitoring the system utilization. In the contrary, the system will show low utilization for its virtual machines thus misleading the admin or the onlooker into believing that the system might be very efficient.
Hence, there is a need of a method and system for method and system for monitoring health of a virtual environment.